We are using self-replicating alphavirus vectors to study the interaction between retroviral envelope and receptor proteins. Cells containing vectors encoding retroviral envelope (or receptor) fuse with cells expressing retroviral receptor (or envelope) and thereby spread the self-replicating vector. This system models the spread of an infectious agent by cell-cell contact. Cells containing these vectors also produce vesicles containing viral envelope or receptor protein on the surface and vector RNA inside. These vesicles act like primitive viruses by fusing with cells bearing the reciprocal protein (envelope or receptor) and transferring vector RNA. This system permits study of the effect of mutations in envelope and/or receptor on virus-mediated fusion and allows genetic methods to be used to select for envelope and receptor variants that facilitate vector spread.Using this alphavirus expression system, we studied various aspects of virus-mediated membrane fusion. We found that the receptor for a model mouse retrovirus (MLV) localizes in cholesterol-rich, plasma membrane regions known as rafts. The receptor binds to another raft-associated cell protein, caveolin, possibly involved in endocytosis. Depleting cells of cholesterol did not affect cell surface expression of the receptor but disrupted rafts and inhibited MLV infection by 95% in mammalian cells and over 99.5% in insect cells which could be more fully depleted of cholesterol. In contrast to the receptor, the virus envelope protein was not raft-associated, and cholesterol depletion of membranes expressing envelope had little effect on fusion with membranes expressing receptor. Several other viruses, including HIV, measles and influenza, are cholesterol-rich or bud-out of, or fuse into, raft-like regions. Cholesterol probably plays a fundamental role in fusion processes since many cell processes that involve membrane fusion, such as endocytosis and vesicle transport, are inhibited by cholesterol depletion. Cholesterol alters membrane properties such as stiffness and curvature that likely affect fusion. Better understanding of the role of cholesterol in MLV fusion should shed light on fusion processes in general.